Nak Kwan Sung. et al : Single-voxel Proton MR Spectroscopy of the Basal Ganglia in Patients with Neurofibromatosis Type 1 internal capsule, cerebellar

Save this PDF as:
 WORD  PNG  TXT  JPG

Size: px
Start display at page:

Download "Nak Kwan Sung. et al : Single-voxel Proton MR Spectroscopy of the Basal Ganglia in Patients with Neurofibromatosis Type 1 internal capsule, cerebellar"

Transcription

1 ? ιj Korean Radiol Soc 1998; 39 : Single-voxel Proton MR Spectroscopy of the Basal Ganglia in Patients with Neurofibromatosis Type 1 1 Nak Kwan Sung, M.D., Jong Ki Kim, Ph.D., Kyu Hyen Oh, M.D., Young Hwan Lee, M.D. Duk Soo Chung, M.D., Ok Dong Kim, M.D., Dong Kuck Lee, M.D. 2, Jin Bok Hwang, M.D. 3 Purpose : To demonstrate the proton MR spectroscopic characteristics of non-neoplastic focal basal ganglia lesions with high signal intensity on long TR MR images in patients with neurofibromatosis type l(nf-l), and to compare them with those of normal-appearing basal ganglia in patients without focallesions. Materials and Methods : Single-voxel proton MR spectroscopy was performed in six patients with NF-l from two families(three with and three without non-neoplastic focal brain lesions). All six individual spectra were obtained from basal ganglia with voxel sizes of about 1 X 1 X 1 cm, three from focal pallidallesions in patients with focal lesions and three from normal-appearing basal ganglia in patients without focal lesions. Spectra were acquired using a 1. 5T clinical MR imager and stimulated echo acquisition mode sequence, with the following parameters : 30 ms of echo time, 13.7 ms of mixing time, and 2560 ms of repetition time. Zero and first-order phase correction was performed. Results : N-acetyl aspartate(naa)/creatine(cr) ratios were similar between focal basal ganglia lesions and normal-appearing basal ganglia, though the former showed slightly lower choline(cho)/cr ratios and slightly higher NAA/Cho ratios than the latter. Relatively enhanced resonances around 3.75 ppm, assigned as glutamate/glutamine, were observed in the spectra of three focallesions. Lipid resonances around slightly different positions were observed in all six patients, regardless ofthe presence or absence offocallesions. Conclusion : Slightly decreased Cho levels and relatively enhanced glutamate/ glutamine resonances are thought to characterize the focal basal ganglia lesions of NF-l. Different mobile lipids appear to be present in the basal ganglia of NF-l patients, regardless ofthe presence offocallesions. Index words: Brain, metabolism Brain, MR Magnetic resonance(mr), spectroscopy Neurofi bromatosis Neurofibromatosis type I(NF-l) or von Recklin ghausen s disease is the most common phacomatosis, 'Department of Diagnostic Radiology, School of I; edicine, Catholic University of TaeguιI-Hyosung 'Department of Neurology, School of Medicine, Catholic University of Taegu-Hyosung Department of Pediatrics, School of Medicine, Catholic University of Taegu-Hyosung Received March 16, 1998; Accepted May 27, 1998 Address reprint requests to: Nak Kwan Sung, M.D., Department of Diagnostic Radiology, School ofmedicine, Catholic University oftaegu-hyosung, ~ , Taemy ung-4dong, Narn -gu, Taegu, , Korea Tel Fax with a prevalence of about one in 3000 births. It is transmitted as an autosomal dominant trait in approximately one half of cases, but spontaneous mutations involving the long arm of chromosome 17 are known to be responsible for the remainder(i). The most common intracranial lesions reported in up to 80 % of patients with NF-l are focal areas of high signal intensity on long repetition time(tr) magnetic resonance(mr) images, and are commonly demonstrated in the basal ganglia, especially the globus pallidus, brain stem,

2 Nak Kwan Sung. et al : Single-voxel Proton MR Spectroscopy of the Basal Ganglia in Patients with Neurofibromatosis Type 1 internal capsule, cerebellar white matter, dentate nucleus, and cerebral white matter( 2-5). These lesions a re generally not associated with vasogenic edema, mass effect, hemorrhage or contrast enhancement, but unusuallesions in the globus pallidus may reveal a mild mass effect or hyperintensity on T1-weighted images (5-8) The purpose of our study was to determine the proton MR spectroscopic characteristics ofnon-neoplastic focal basal ganglia lesions in patients with NF-1 using stimulated echo acquisition mode (STEAM) sequence with small voxel sizes of about 1 X 1 X 1 cm, and to compare them with those of normal-appearing basal ganglia in patients with no focallesions. Materials& Methods Six patients from two families(two girls and their father from one family, and two boys and their father from another family ; patient age range, years; mean age, 24 years) were diagnosed as NF-1 on the basis of the criteria of the National Institute of Health (NIH) Consensus Development Conference. Three of the six patients showed typical multifocal brain parenchymallesions on one or both sides of the basal ganglia, thalamus, internal capsule, cerebral peduncles and dentate nucleus, but in the remaining three the brain appeared normal on MR images. The six basal ganglia lesions (bilateral abnormalities in three patients with focal brain lesions) measured between 0.9 and 1.5 cm in their greatest dimension. Normal variations are known to be present in metabolite concentrations at different anatomic sites, and to improve the reliability of sampling for single-voxel proton MR spectroscopy all six individual samples were thus obtained from the same site. Three spectra (patient age range, years; mean age, 11 years) were obtained from the focal lesions of basal ganglia and three (patient age range, years ; mean age, 37 years) from the regions of basal ganglia of normal appearance in which MR images did not reveal focal lesions. In patients with bilateral basal ganglia lesions, the larger one was selected for sampling. Spectral voxel was selected graphically from axial T2-weighted scout images with a voxel size of about 1 X 1 X 1 cm(fig. 1A). In the case of a smalllesion measuring about 0.9 cm in its greatest dimension, an even smaller voxel size was selected ; this was so as to minimize the contamination Cr Fig. 1. A 13-year-old male NF-1 patient with focal brain lesions on MR lmages. A. Axial T2-weighted scout image shows the location of MR spectroscopic single-voxel (about 1 X 1 X 1 cm) at the left pallidal lesion. There also showed right pallidal lesion at the lower l eve 리래 l(no 야 t shown 미 1 associated bilateral thalamic lesions at the pulvinars. B. Proton MR spectra show slight de- A B crease üf Cho level at 3.22 ppm, relative enhancement ofthe glutamate/glutamine(glx} resonances around 3.75 ppm, and observable lipid resonances around 1.05 ppm. NAA Li p Cr ~ 3 NAA Fig. 2. A 42-year-old male NF-1 patient without focal brain lesions A. Axial T2-weighted scout image shows the location of MR spectroscopic single-voxel (about 1 X 1 X 1 cm) at the normal-appearing right basal ganglia. B. Proton MR spectra show observable lipid resonances around 1. 5 ppm without particular alteration of the NAA and Cho levels A B m

3 끽J Korean Radiol Soc 1998; 39 : of surrounding normal brain. In ca,ses in which brain appeared normal on MR images, we centered the voxel to the basal ganglia(fig., 2A). Shimming of selected voxel was performed until a water line width of 3 Hz was obtained. Water suppression was achieved by the prior application of chemical-shift selective (CHESS) three orthogonal radiofrequency pulses. All proton spectra were acquired using a 1.5 T GE Signa Horizon clinical MR imager and STEAM sequence for localization, with the following parameters: 30 ms of echo time, 13.7 ms of mixing time, and 2560 ms of repetition time. Spectral width was 2000 Hz, and was stored in 2048 data points. The typical average number of scans was Raw data were zero-filled to 4096 and exponentially apodized with 1. 5 Hz line broadening, and subsequently Fourier-transformed. Zero and firstorder phase correction was performed, though there was no baseline correction, further smoothing or resolution enhancement. Results NAA/Cr, Cho/Cr, and NAA/Cho ratios of the three basal ganglia lesions were 1.07 ::!::: 0.10, 0.46 ::!::: 0.10, and 2.46 ::!::: 0.72, respectively, whereas these ratios in the three regions in which basal ganglia appeared normal with no focallesions, were 1.08 ::!::: 0.08, 0.73 ::!::: O. 08, and 1.48 ::!::: 0.08, respectively(table 1). Focal lesions showed slightly lower Cho/Cr ratios and slightly higher NAA/Cho ratios than brain regions of normal appearance, but NAA/Cr ratios were similar. Analysis ofproton MR spectra observed from focal basal ganglia lesions therefore showed no distinctive change in NAA levels but slightly lower Cho levels than those obser ved in basal ganglia ofnormal appearance. Relatively enhanced resonances of around 3.75 Table 1. Metabolic Ratios Between Focal Basal Ganglia Lesions and Normal-appearing Basal Ganglia in Patients withnf-l Basal ganglia lesions Case 1.11/F Case 2.10/F Case 3.13/M mean ::1: SD Normal-appearing B.G Case 4.16 /M 1.02 Case 5.42/M 1.05 Case 6.52/M 1.17 NAA/Cr Cho/Cr NAA /Cho ::1: ::1: ::1: mean ::1: SD 1.08 ::1: ::1: ::1: 0.08 m ppm, assigned as glutamate/glutamine, were observed in the spectra of three focal lesions(fig. 1B). Lipid resonances around slightly different positions were observed in allsix patients with NF-1, regardless ofthe presence or absence offocallesions(fig. 1B, 2B). Discussion Patients with NF-1 exhibit a wide variety of neoplastic and non-neoplastic intracraniallesions(1). The most common abnormality demonstrated by neuro-imaging studies is non-neoplastic high-signal-intensity foci in the basal ganglia, especially in the globus pallidus, brain stem, internal capsule, cerebellar white matter, dentatenucleus, and cerebral white matter, as seen on long TR MR images(2-5). Because no correlation has been established between pathologic and radiologic findings, the high-signal-intensity foci seen on long TR MR images have confusingly been described as hamartomas(5), heterotopias( 4, 6), gliosis(6), low grade tumors(3), or disordered myelination(8). Follow-up MR imaging studies demonstrated critical evidence oftemporal evolution ofthe lesions, and this transient nature with spontaneous regression would thus deny the possibility of developmental abnormalities such as hamartoma, heterotopia, or neoplastic proliferation( 8, 9). The fact that the lesions seen on MR images were found only in children aged less than 13(mean age, 11 years), but not in older children or adults(mean age, 37 years) supports the hypothesis of evolution of focal brain lesions in patients with NF- 1. A recent histopathologic study by Di Paolo et a1(10) suggested that hyperintense foci on T2-weighted images were related to spongiform myelinopathy or vacuolar change (vacuoles ranging from 5 to 100 띠 n in diameter) of myelin which appeared to constitute the predominant pathologic change. There was no evidence of inflammatory reaction or demyelination. They believed that water within the vacuoles was responsible for the occurrence of high signal intensity on T2-weighted images, and that isointensity rather than hypointensity, as seen on T1-weighted images, was due to partial volume averaging of water within small vacuoles and intervening tissue. Microcalcifications and perivascular schwannosis seen in the globus pallidus might account for the unusual hyperintensity seen on T1-weighted images. Using large voxel sizes measuring 2 X 2 X 2 cm to 3 X 3 X 3 cm, Castillo et a1(ll) reported proton MR spectroscopic characteristics of the brain in patients with NF-1 and showed that focallesions had proton spectroscopic patterns similar to those of normal brain; this might,

4 Nak Kwan Sung. et al : Single-voxel Proton MR Spectroscopy of the Basal Ganglia in Patients with Neurofibromatosis Type 1 we think, influence their resu1ts, derived as they were from partia1 vo1ume averaging of norma1 surrounding brain within the vo1ume samp1ed. The 10ng-echo-time point reso1ved spectroscopy (PRESS) sequence they employed is subject to limitation when used to detect resonances from a coupled spin system such as glutamate, glutamine or inositol. We used the STEAM sequence to revea1 any subtle changes in more metabolites that have short decay times In our antegrade study, all samp1es for sing1e-voxel proton MR spectroscopy were obtained from the same anatomic sites ; this was because metabolite concentrations at different anatomic sites are known to vary( 12) Basal ganglia were selected for sampling because they have been reported to be the most frequent site offoca1 1esions, which are 1arger than those found e1sewhere So as to include as much of foca1 basa1 ganglionic lesions or regions of norma1-appearing basa1 ganglia in patients without foca1 1esions as possib1e, while minimizing partia1 vo1ume averaging, optima1 sizes and 10cations ofvoxels were chosen. Our proton MR spectroscopic stud y of the brain in patients with NF-l revea1ed that NAA 1eve1s observed in foca1 basal ganglia lesions were not significant1y different from those in norma1-appearing basa1 ganglia without foca11esions or from those previous1y reported in norma1vo1unteers. NAA is mostly found in neuronal cells and some neuroglia1 precursors such as 0-2A progenitor cellline, and is thus accepted as a neurona1 marker. Norma1 leve1s of NAA in focal brain parench ymallesions in NF-l suggest that there is no particu1ar 10ss or damage of neurona1 cells and/or axona1 processes of the 1esions, and this wou1d exp1ain the transient nature of the 1esion and its spontaneous regression without residua1 abnorma1 signa1 on followup MR images. Proton MR spectroscopy is therefore a usefu1 diagnostic tool for the differentiation of a focal lesion from a glioma, another common CNS 1esion in patients with NF-l(ll). Cho 1eve1s appeared to be slightly lower in focal NFl-type brain 1esions. The in vivo Cho signa1 consists of various compounds such as choline, phosphocho1ine, glycerophosphocho1ine, acetylcholine and betaine, all of which are components of phospholipid metabo1ism and constituents of cell membranes in the brain. Thus, decreased Cho signa1 intensity in foca11esions may reflect some change other than demyelina in foca1 basa1 ganglia lesions of NF-l, but not in the spectra ofnorma1-appearing basa1 ganglia. In addition, lipids resonating around slightly different positions were observed in al1 six patients, and this may represent the presence of different mobile lipids in both foca1 1esions and regions of normal-appearing brain without focal1esions. Changes in glutamate/g1utamine and lipid metabolism ofthe brain in patients with NF-l have not been demonstrated by previous proton spectroscopic studies; this is because the researchers invo1- ved used the PRESS sequence, which is inherently insensitive to glutamate/g1utamine and lipid(ll). In man, in vivo spectroscopic studies of cerebra1 glutamate/ glutamine concentrations have been performed main1y in chronic hepatic encephalopathy patients, in whom it was postu1ated that glutamine acted as an osmo1yte (13). The cause ofincreased glutamate/g1utamine 1eve1s on1y in foca1 basal ganglia lesions and the appearance of 1ipid resonances regardless of the presence or absence of foca1 1esions in patients with NF-l is unclear and requires further investigation. In summary, the spectra of presumptive non-neop1astic focal basa1 ganglia 1esions in patients with NF-l are characterized by slightly decreased 1eve1s of Cho and re1ative1y enhanced glutamate/g1utamine resonances without particular changes in NAA levels. References 1. Osborn AG. Diagnostic neuroradiology. 1st ed. St. Louis: Mosby, 1994: Brown EW, Riccardi VM, Mawad M, Handel S, Goldman A, Bryan RN. MR imaging of optic pathways in patients with neurofibromatosis. AJNR 1987; 8: Hurst RW, Newman SA, Cail WS. Multifoca l intracranial MR abnormalities in neurofibromatosis. AJNR 1988; 9 : Bognanno JR, Edwards MK, Lee TA, Dunn DW, Roots KL, Klatte EC. Cranial MR imaging in neurofibromatosis. A JNR 1988 ; Aoki S, Barkovich AJ, Nishimura K, et al. Neurofibromatosis type 1 and 2: cranial MR imaging. Radiology 1989; Mirowitz SA, Sartor K. Gado M. High-intensity basal ganglia lesions on Tl-weighted MR images in neurofibromatosis. AJR 1990 ; 154: Pont MS, Elster A. Lesions of skin and brain: modern imaging of the neurocutaneous syndrome. AJR 1992; 158: Sevick RJ, Barkovich AJ, Edwards MSB, Koch T, Berg B, Lempert T. Evolution of white matter lesions in neurofibromatosis type 1: MR findings. AJR 1992 ; 159: Itoh T, Magnaldi 5, White RM, et a1. Neurofibromatosis type I the evolution of deep gray and white matter MR abnormalities. AJNR 1994;15: DiPaolo DP, Zimmerman RA, Rorke LB, Zackai EH, Bilaniuk LT, Yachnis AT. Neurofibromatosis type 1: Pathologic substrate of high-signal-intensity foci in the brain. Radiology 1995; 195:

5 J Korean Radiol Soc 1998; 39: Castillo M, Green C, Kwock L, et aj. Proton MR spectroscopy Radiology 1996; 199: in patients with neurofibromatosis type 1 evaluation of harn- 13. Takahashi H, Koeh ler RC, Brusilow SL, Traystman RD. lnhibiartomas and clinical correlation. AJNR 1995; 16: tion of brain glutarnine accumulation prevents cerebral ederna 12. Lu D, Pavlakis SG, Frank Y, et aj. Proton MR spectroscopy of in hyperarnmonemic rats. Am J Physiol 1991; 261: the basal ganglia in healthy children and children with AIDS 대한방시선의학호 ');'1998;39: 신경섬유종증 1 형환자에서기저핵의단일화적소양성자자기공명분광소견 1 1 대구효성가톨릭대학교의과대학진단방사선과학교실 2 대구효성가톨릭대학교의과대학신경과학교실 3 대구효성가톨릭대학교의과대학소아과학교설성낙관 김종기 오규현 이영환 청덕수 검옥동 이동국 2. 황진복3 목 적 : 신경섬유종증 1 형환자에서긴반복시간자기공명영상에고신호강도로보이는기저핵의비종양성국 소병변의양성자자기공명분광검사상특정을찾아보고, 국소뇌병변이없는신경섬유종증 l 형환자에서정상으 로보이는기저핵의자기공명분광검사결과와비교하고자하였다. 대상및방법 :2 가족 6 명의신경섬유종증 1 형환자에서단일화적소양성자자기공명분광검사를실시하였고, 이중 3 명은비종양성국소병변이있었고 3 명은없었다.6 개의모든표본은약 lx l xlcm 크기의화적소를사용하여기저핵에서얻었고, 이중 3 개는국소뇌병변이기저핵에있었던환자의담창구병변에서그리고나머지 3 개는국소뇌병변이없었던환자의정상으로보이는기저핵에서얻었다. 1.5 Tesla 자기공명영상장치를사용하 였고, 30msec 의에코시간, msec 의혼합시간그리고 2560msec 의반복시간으로자극에코획득방식을이용하 여스펙트럼을얻었다. 영점그리고일차위상수정을실시하였다. 결과 : N-acetyl aspartate(naa)/creatine(cr ) 비는기저핵의국소병변과정상으로보이는기저핵에서 비슷하였으나, 전자가후자보다 choline ( Cho) / C r 비가조금낮고 NAA/Cho 비는조금높았다. Glutamate/ g l utamine에해당하는 3. 75ppm근처에서비교적증가된공명이 3예의국소병변스펙트럼에서관찰되었고, 서로 약간다른위치에서공명하는지방은국소병변의동반유무와관계없이 6 예모두에서관찰되었다. 결 론 : 약간감소된 Cho 치그리고비교적증가된 Glu tamate / G lutam i ne 공명이신경섬유종증 1 형에서보이 는기저핵의국소병변의특정으로생각된다. 그리고상이한유통성지방들이국소병변의유무에관계없이 NF -I 환자의기저핵에존재하는것으로생각된다

6 대한방사선의학회쩌 154 자학술대회및종회사전등록안내 다일시 : 1998 년 10 월 15 일 ( 목 ) - 17 일 ( 토 ) n 장 소 : 호텔롯데월드 口명점 : 15평점口샤전등록 : 예년과같이학술대회의원활한운영을위해사전등록제를시행합니다사전등록을하신회원에게는학술대회초록집, 연수교육교재등을사전에우송하여드립니다. l 등록비 구 분 정 회 원 전공의회원 비 회 원 사전등록마감 (1998년 9월 15일까재 70, 000원 30, 애0원 70, 때0원 현장등록 (1998년 9월 15일이후 ) 80, 애0원 40, 어0원 80, 애0원 l 범주별연쇄정 (Categorical Cou 연 e) 교삐 구분 사전등록마감 (1998 년 9 월 15 일까지 ) 정회원 l 권당 5, 때원전공의회원 권당 5, 애 0 원비회원 1 권당 5, 000 원 현장등록 (1998 년멜 15 일이후 ) 권당 7, 애0원 권당 7, 애0원 권당 7, 어0원 l 시전등록 t병사전등록은아래구좌로송금이완료된경우만인정하며, 송금한후학회 Home Page를이용하여신청하시거나, 별지의사전등록신청서를작성하신후신청서를우편또는 Fax로송부하시면됩니다. 거래은행 : 평화은행구좌번호 : IDl5-373 예금주 : 대한방사선의학회 I {Y 성빙법 [j] Home Page 를이용하는방법 학회 HomePage의 URL 주소인 http ://radiol. me빠mrea, net으로접속핸후추계학회 Site로가서등록하끄 } 하는 내용을입력하시면됩니다. 전송이완료되면즉시접수번호를부여받습니다. 낌 을이용하는방법 학회 주소인 chollian.net 으로송신하시면되며, 귀하의 주소로다음날답신이됩니다. 디초록제출 초록제출미감 : 1998 년 7 월 31 일 ( 금 ) 초록제출빙법 : 사전등록신청방법과동일 디획 의 조를 일시 자 소 대한방사선의학회총회 10월 16일 12:00-13:30 대한X선검진협회총회 10월 15일 11 :30-13:00 정기이사회 hirebroom( 지하1 층 ) χω이ι